Sound is captured by microphones and stored in computer sound files which may be manipulated and altered by software algorithms for the purposes of removing noise defects, such as but not limited to hums, buzzes, pops, clicks, snaps, hisses, wireless ringing, traffic, and engines. The human brain has a subconscious ability to automatically remove these noise defects from consciously being heard, however in sound recordings these noise defects become much more apparent and must be removed through the use of software.
Depending on the environment and equipment used, microphones simultaneously capture desired sounds and undesired noise. For example, a wireless microphone worn by an actor on a production set may also capture noise defects from various electromagnetic or environmental sources, on or off the production set. A common example is a power generator or air conditioner humming in the distance. Noises in a sound recording are detrimental to the presentation of an actor's performance and may distract the viewer from the story. An example of this would be hearing a power generator or air conditioner on a deserted island.
Most actors, producers, and directors prefer original performances, and ask professionals to use manual sound processing software to repair the actor's original sound recordings. In the event the noise defects cannot be removed, the actor will have to re-perform their dialog in a studio and duplicate the emotions and feelings of their original performance. This is generally considered difficult, costly, time consuming, and is usually inferior to the original performance.
The manual removal of noise from sound recordings requires years of practice and experience in order to achieve results that are acceptable to professionals. Manual removal of noise also requires extraordinarily repetitive and time consuming manual tasks that are better suited to automation.
Accordingly, it is an aspect of the present invention to provide a system and method for automatically detecting and removing noise defects from sound recordings while preserving the human voice or other desired sounds in the sound recordings. This invention automatically and accurately removes obvious noise defects and mitigates manual repetitive labor.